Problems With Holding the Altitude

[gitbatu]

I have two issues I'd like to open to discussion:

1- Our quadcopter wasn't able to hold its altitude during missions. It was supposed to hold at a waypoint for 30 seconds, but instead, it descended slowly and jumped repeatedly during that period. What might be causing this? (running Px4 firmware 1.8.2 on a generic quad x with Pixhawk 1. 3DR GPS module connected.) Log: https://review.px4.io/plot_app?log=9ace9200-b9e7-4b3a-9090-4720aac6dd74

2- Another quadcopter rocketed to the sky at the start of a mission. We panicked and pulled it down with a rope we had attached earlier. The mission was to takeoff to 4 meters and hold there for 10 seconds, then land. This one has the same hardware as the other except that it has a ReadyToSky Ublox M8N GPS module and a customized airframe. Why do you think that happened? Log: https://review.px4.io/plot_app?log=83478e75-9a39-4197-a9f4-8fc9ad0e889c

After these incidents, I have tried the same mission in jMAVSim with HITL simulation. Turned out the quad could not hold its altitude in there as well. I've downgraded the firmware to 1.7.0 to test if the same thing would happen again. It did. Furthermore, I've observed that the quad loses its global position lock in the simulation sometimes.

I would really appreciate some insight to these problems. Thank you!

[MaEtUgR]

1. Your vehicle detected multiple times it is landed and not only ground contact but all the three landing stages:

I tried to find out why that should be: The vehicle seems to have a good thrust to weight ratio but you adjusted the hover thrust to 0.24 while the minimum thrust is 0.08. I'd have to check the code from 1.8 to check further. It might be different on current master and hence 1.9 beta.

[gitbatu]

Indeed the vehicle detects false landings. Furthermore in land mode, it cannot decide if it actually landed; hence, it touches the ground and jumps a bit and do these repeatedly. Can this be related to hover thrust? I thought it only affected manual mode. I will try it in SITL simulation. (Edit: Vehicle is getting affected by hover thrust value in auto modes as well. But it still can hold its altitude even with a low hover thrust setting.)

Another strange thing is that we see Z velocity setpoint going up and down. I would expect a stable setpoint:

Can we interpret this as an altitude controller issue maybe?

By the way, may I ask how you obtain that plot? Whenever I open a log in FlightPlot I fail to plot anything. Is there a trick to using it?

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[stale]

Closing as stale.